Fullscreen HTML5Med Res (??MB)HTML4

REMOTE CONTROL BY BOB YOUNG An 8-channel decoder for radio control This decoder is designed to mate with the AM receiver described in the previous four months. The PC board is exactly the same size as for the receiver & the two plug into each other so that no interconnecting wires are required. The development of this decoder has been a classic example of the problems thrown up by component manufacturers constantly changing their components. There should have been no difficulty whatsoever in changing my original two-IC design to a surface mount unit, or at least so I thought. I expected to produce two prototypes in the development schedule. What an optimist! To begin with, 74C series ICs are not readily available in surface mount although we did manage to locate some in the USA at about US$3.50 per IC. So I blithely proceeded to substitute 74HC series ICs which are available over the counter at several large component stores. All hell broke loose. When switched on, the circuit which has been in more or less continuous production from 1974 did not work at all. It needed a lot of work to sort it out. For many years and particularly since the introduction of the very large quarter scale models, there have been mutterings about noise or interference problems related to the long servo leads in these models. The talk was always vague and no one appeared to have any definite idea as to what was the nature of the noise or where it came from. The implication seemed to be that RF was being picked up on these long leads from other transmitters and was then finding its way back into the receiver – much the same as CB transmitters break into older stereo sets. As a result, we were often asked to fit ferrite beads and all sorts of suppressors to long servo leads. I might add here that I spent hours examining my sets and never located any definite signs of this problem. When I finally managed to trick the first prototype decoder into working, the very first thing I noticed when I plugged in a servo was a very strong noise spike at the receiver detector, associated with any channel which had a servo lead attached. Removing This photo shows how the 8-channel decoder sits in the bottom of the case & the receiver plugs into and sits on top of it. Note the slot in the decoder board to give access to the crystal on the receiver board. 70  Silicon Chip April 1995  71 B E VIEWED FROM ABOVE C E Q1 BC848 C C11 B .01 R12 10k C9 .001 R11 47k R16 1M 1 7 14 C12 xx 2 IC2a 40106 3 D2 BA516 IC2b 4 SILVERTONE MK22 8-CHANNEL DECODER R15 100k R18 1M C15 1.5 C16 0.1 5 6 C13 1.5 C10 .033 R9 1k IC2c R10 100k Fig.1: the decoder takes the serial data stream from the receiver & produces up to eight pulse outputs to drive the servos. IC2 is essentially a pulse shaper, while IC1 is the shift register where the decoding actually takes place. RX IN TB10 R14 100  C14 47 +4.8V R13 220k D1 BA516 O2 O1 O0 5 4 3 EXPANSION TB9 7 R1 1k 6 O3 10 IC1 O4 1 74HC164 11 O5 A 12 2 B O6 13 O7 9 MR 8 CLK 14 R2 1k R3 1k R4 1k R5 1k R6 1k C1 .001 C2 .001 C3 .001 C4 .001 C5 .001 C6 .001 C7 .001 R7 1k C8 .001 R8 1k R17 56  CHANNEL 8 TB1 CHANNEL 7 TB2 CHANNEL 6 TB3 CHANNEL 5 TB4 CHANNEL 4 TB5 CHANNEL 3 TB6 CHANNEL 2 TB7 CHANNEL 1 TB8 Fig.2: this is a typical data stream from the receiver, as meas­ured at the collector of Q6. Fig.3: this is the same data stream as in Fig.2 after it has been squared up by IC2a. Fig.4: this is output waveform from IC2b, showing the synchronisation pulse. Fig.5: this is typical of the pulse output that will be found on any of the servo lines from IC1. the servo lead caused the spike to disappear. It was fairly obvious that the high speed switching (about 15MHz) was radiating from the servo lead. Is this the problem that modellers were concerned about? From memory it was about the time of the introduction of high speed CMOS that the noise was first mentioned. The problem was however, what was I going to do about it? CMOS surface mount was not available and I had already gone into print and promised 8, 16 and 24-channel decoders. (4000-series CMOS is readily available in surface mount but there is not a suitable 8-bit shift register in this series). So here I was with a decoder that did not work reliably and when it did, it radiated like a transmitter. It was while discussing these prob72  Silicon Chip lems with a colleague that the answer to the entire dilemma popped up. My friend showed me an article in an electronics magazine which stated that HCMOS chips ring like bells in the output stage and that an anti-ringing filter was most helpful, especially on clock lines. This article went on to say that a 1kΩ resistor followed by a 1000pF capacitor was all that was required to cure the problem. The circuit diagram of Fig.1 shows the arrangement. The addition of the filter in the servo leads eliminated the radiation completely and the decoder began working reliably when the filter was placed in the clock line between IC2 and IC1. However, I am really annoyed about this whole affair. In the case of the 24-channel decoder, I am now stuck with adding 51 components on PC boards that are too small to accom­modate this many components – all this to get rid of switching speed I do not need. In the end, the 8-channel decoder called for a compromise and I used a 40106 in place of the 74HC04 (unfortunately, I could not change the 74HC164). This at least got rid of the filter on the clock line and I managed to complete the PC board layout without jumpers and with all components in place except for C1 which ended up on the bottom layer. I was not so lucky with the 16-channel expansion PC board, unfortunately. Here I ended up with about six jumpers. This module will be presented next month and features a double sided surface mount board. Still, the completed 24-channel receiver is a very professional looking piece of TB1 TB2 TB3 TB4 TB5 TB6 TB7 TB8 TB9 C10 R14 R18 C15 R15 C11 C9 R11 R9 R10 R12 C12 R13 C14 R2 C2 R3 C3 C4 R4 C5 R5 C6 R6 C7 R7 C8 R8 R16 R17 IC2 40106 R1 (IC2) is used as a pulse shaper and driver for the 74HC164 shift register (IC1). Inverter IC2a provides the clock data C1 (as shown in the scope photo of Fig.3) and also drives IC2b. D1 D2 1 IC2b’s output supplies the IC1 synchronisation pulse (shown 74HC106 in the scope photo of Fig.4) in Q1 association with D2, R10 and 1 C10. During the long pause TB10 C14 TB10 between pulse frames (6ms C16 minimum), C10 charges via R10 and lets pins 1 and 2 on IC1 go Fig.6: here are the component overlays for the top & bottom of the 8-channel decoder high, ready for the first pulse on board. Only a single capacitor (C1) & the 3-way header are mounted on the underside (see text). the next frame. R9 is included to introduce a small delay in work. All of the PC boards simply supply decoupling network for the the switching, to stop mistriggering. plug together. receiver and decoder. The signal pin IC2b also drives IC2c which develFinally, I have just a few words on on TB10 goes to the audio slicer which ops a chip enable voltage at pin 9 on the servo leads them­ selves. One of consists of Q1, C15, R18, R15 and IC1. This acts as a fail-safe in the abthe problems faced by modellers with R12. The input floats on the receiver sence of the incoming pulse train and older equip­ment is the need for re- noise floor and rejects the bottom 1V thus helps to stop servo gears being placement receivers. The transmitters of hash. Thus only clean high level damaged. C13 and R13 smooth out never seem to wear out and servos are audio pulses are fed to the audio the pulses and provide approx­imately fairly robust but receivers often die and amplifier. The scope photo of Fig.2 +4.5V DC on pin 9, thus enabling the the agents often discontinue service on shows the signal from the receiver (at chip. Loss of signal sends pin 9 low, the collector of Q6). older models. shutting down IC1 and completely C11, R11 and C9 form a filter to re- eliminating spurious outputs on the This leaves the modeller with an unuseable system. Added to this is the move any remaining hash. The 40106 servo lines. confusion brought about by non-standardisation of the servo plugs. Most servos these days plug into header pins Receiver & Decoder Kit Availability mounted directly onto the PC board but the arrangement of these header Receiver PC board (double-sided with plated-through holes) ..........$11.50 pins can vary from manufacturer to Basic receiver kit: all parts except crystal .........................................$45.00 manufacturer. Built & tested AM receiver less crystal .............................................$59.00 This new receiver/decoder package is designed to replace as wide a Decoder PC board (double-sided with plated-through holes) ..........$11.50 variety of receivers as possible and a 8-channel decoder kit: all parts less servo pins or connec­tors .........$32.00 considerable amount of thought has Built & tested 8-channel decoder but less servo plugs ....................$45.00 gone into making this possible. To Expansion kit: all components to build the 16-channel decoder ......$42.00 begin with, the polarity of the power pins may be reversed by simply Built & tested 16-channel decoder less servo connectors ...............$55.00 cutting two tracks and jumpering. 8-channel receiver case (includes labels) ........................................$11.50 In addition, the header pins may be 16-channel receiver (includes labels) ...............................................$19.50 replaced with fly leads for even more Machine wound RF coils ....................................................................$2.95 versatility. Machine wound IF coils ......................................................................$2.95 Circuit operation Crystals (AM) per pair ......................................................................$17.95 The decoder is contained on a sepServo header pins (each) ...................................................................$0.12 arate PC board and con­nects to the receiver through a 4-pin header plug Futabe EXT lead .................................................................................$3.40 (TB10). Power to the receiver is deJ.R. EXT lead ......................................................................................$3.40 rived from the power rails associated Sanwa EXT lead .................................................................................$3.40 with the servo plugs. Depending upon the number of channels in use, you can either use a spare servo output as the power input or if all eight channels are in use, a “Y” or splitter lead can be inserted between one servo and header pins. R17, R14, C14 and C16 form a Notes: (1). When ordering crystals, do not forget to specify frequency. (2). All orders should add $3.00 for postage and packing. Payments may be made by cheque, money order, Bankcard, Visa Card or Mas­tercard. Send all orders to Silvertone Electronics, PO Box 580, Riverwood, NSW 2210. Phone (02) 533 3517. April 1995  73 Provided the conditions are all correct on pins 1, 2, 8 and 9, the pulses will clock through the shift register and servo outputs will appear at pins 3, 4, 5, 6, 10, 11, 12 and 13, as shown in the scope photo of Fig.4 (ie, if all eight pulses in a frame are transmitted). If only two pulses per frame are trans­ mitted, then output 3 will be the sync pause and output 4 will be channel 1 again and output 5 will be channel 2; output 6 will be the sync pause and so on. Thus, in a 24-channel receiver channel 1 will appear three times if only eight pulses are transmitted. This is a useful feature during testing if only transmitters with a lesser number of channels are available or it can be very useful as a splitter/driver for parallel servo operation. In this case, each output only drives one servo as against two in the case of a “Y” lead. The three unused inputs on IC2 (pins 9, 11 & 13) are tied to ground. Finally, TB9 is the expansion port for the 16-channel add-on PC board. This port carries clock, data and enable infor­mation, as well as the two power rails. Construction The PC board provided with the kit is a double sided plated-through board with solder resist over all but the component pads. For those not familiar with surface mount assembly, read the article on this subject in the January 1995 issue of SILICON CHIP. The component overlays for the top and bottom of the boards are shown in the diagrams of Fig.6. First, the polarisation of the power rails must be decided and set accordingly. As delivered, the PC board is set up for centre rail positive (JR, Futaba, Hi Tech). To reverse this order (KO, Sanwa), simply cut the thin tracks connecting the power rails with the decoder supply rails (along the top edge of the board as shown in Fig.6) and reconnect them to the appropriate rails. There are pads located alongside the power rails for this purpose. Note that one track is located on the top layer and the other on the bottom layer. Use 10amp fuse wire or a component lead for the jumper. No reverse voltage protection Be very careful here for there is insufficient voltage for a reverse voltage protection diode when using a 4.8V 74  Silicon Chip Fig.7: this exploded diagram shows how the decoder & receiver sit in the case. The various slots in the case give access to the crystal & provide exit holes for the antenna & servo lines. battery. Whilst on this subject, the receiver is set up for 4.8V and will not operate satisfactorily from a 6V battery. If you need to operate from 6V then insert two diodes in series with the +6V lead, to reduce the voltage by 1.2V. Be certain to mark your finished unit clearly because if you end up with two receivers, one positive and one negative, you could land yourself in bother at some later date. Begin assembly by mounting the SM devices and solder one pad on each component first. Order is not important here, just suit yourself. Once all of the SM components are mount­ed, mount the two capacitors. C14, the 47µF tantalum, is polarised so be careful to follow the markings. Next, mount the 3-pin socket (TB10), making sure that it is on the correct side of the PC board. This is on the opposite side to the components. If the thought of having a plug in the systems worries you in regard to vibration, then this connector pair may be deleted and replaced with wire connections. At this point, it needs to be clearly understood how many channels will be required and whether fly leads or pins are to be used for the servo connectors. Presumably you have ordered a kit and specified the number and type of servo connectors required. If you are using fly leads, just solder the leads into the appro­priate holes in the servo connector pads in the order they lay on the servo lead. If the leads are centre-negative, do not forget to reverse the PC board connections if you have not already done so. If you do decide to use fly-leads for the servo outputs, you will need to file one or two slots in the case end for the lead exits. Do not forget to thread the servo leads through the grommets before soldering them to the PC board (see the exploded case diagram of Fig.7 for details). If you intend to use the pins, then just simply push the 3-pin plug through the PC board with the plastic base on the component side and with the long pins going through the holes. Solder the pins from the reverse side. Snip off the excess pins on the reverse side and remove the plastic from the pins on the component side. This now leaves pins the correct length for a servo socket on the component side of the PC board. If you intend using more than eight channels, you must now install the expansion port. Follow the same routine as for the servo pins. You now have a finished decoder. Testing Plug the decoder into a pre-tuned receiver and leave both units out of the case. It is wise to insert a piece of insu- lating card between the two boards, as otherwise they can touch if bumped. Once they are snapped into the slots in the case, this is not necessary. Testing can now proceed as all components are accessible from the servo pin side of the PC board. Alternative­ly, an extension lead can be made up to keep the two PC boards well separated during servicing and testing. Turn on the associated transmitter and, using an oscillo­scope, check the input to the slicer and compare the waveshape with Fig.2. Next proceed to check pins 1-6 on IC2. These should compare with Fig.3 on the odd-numbered pins and should be inverted on the even-numbered pins. Now test IC1 pins 1 and 2 and compare the waveshape here with Fig.4. Pin 8 on IC1 compares with Fig.3 and pin 9 should be a DC voltage with a low level of ripple on top floating at about +4.5V above ground. Whilst monitoring pin 9, switch off the transmitter and note that it goes low no more than one second after switch off. If all of the foregoing is in order, the output at pins 3, 4, 5, 6, 10, 11, 12 & 13 will look like Fig.5. Plug in one or more servos and check the operation from the transmitter. Be careful not to reverse the servo plug as the polarising key is in the case. Case assembly If you are using the header pin layout, complete the assembly by simply snapping the decoder PC board into the case, with the pins pointing towards the punched holes in the case bottom. Next, plug the receiver board into the 3-pin socket, leaving the fourth pin (closest to the edge of the PC board) outside the socket. This now provides a useful test point to attach an oscilloscope or meter. The receiver simply rests in the notch in the case sides. Slip on the case lid, attach the labels and open the servo slots in the bottom label that you wish to use. Leave any of the unused slots covered to prevent ingress of dust. Secure the lid with a wrap of clear tape. Now go and have some fun. Troubleshooting Now for the sad cases, it is back to the test bench. First­, check the assembly for missing components, soldering faults, etc. Check the decoder power rails to see if they are compatible with the servo leads you are using. Be sure that these have not been accidentally reversed and do not suit the servo leads you are using. Now grab your multimeter and start testing voltages. The input voltage at the power rails will be the battery voltage unless dropping diodes are installed. Next, check the power rails in the decoder. With a nicad battery reading 5.0V, pin 14 on both chips should be approximately 4.9V. Pin 7 on both chips is the ground pin. The base of Q1 should be +0.35V and the collector +4.9V, with no signal from the transmitter. The rest is routine servicing. If all of the DC and input voltages are correct, then you may suspect a faulty IC, but let me tell you, it is rarely ever the IC. I have found from experience that 99 times out of 100, it is an associated fault. If all your best efforts are to no avail, then send it back to Silvertone and we will sort it out for you. Next month, we will describe the 16-channel decoder board. This will be a double-side board with surface mount components on both sides. SC 20MHz Dual Trace Scope $795 100MHz Kikusui 5-Channel, 12-Trace 50MHz Dual trace Scope $1300 COS6100M Oscilloscope $990 These excellent units are the best value “near brand new” scopes we have ever offered. In fact, we are so confident that you’ll be happy, we will give you a 7-day right of refusal. Only Macservice can offer such a great deal on this oscilloscope . . . and you are the winners! 1. Power switch 2. LED 3. Graticule illumination switch 4. Trace rotation 5. Trace focus 6. Trace intensity for B sweep mode 7. Brightness control for spot/trace 8. Trace position 9/10/11. Select input coupling & sensitivity of CH3 12. Vertical input terminal for CH3 13. AC-GND-DC switch for selecting connection mode 14. Vertical input terminal for CH2 15/22. Fine adjustment of sensitivity 16/23. Select vertical axis sensitivity 17/24. Vertical positioning control 18/25/38. Uncal lamp 19. Internal trigger source CH1,CH2,CH3,ALT 20. AC-GND-DC switch for selecting connection mode 21. Vertical input terminal for CH1 26. Select vertical axis operation 27. Bezel 28. Blue filter 29. Display selects A & B sweep mode 30. Selects auto/norm/single sweep modes 31. Holdoff time adjustment 32/51. Trigger level adjustment 33/50. Triggering slope 34/49. Select coupling mode AC/HF REJ/LF REJ/DC 35. Select trigger signal source Int/Line/Ext/Ext÷10 MACSERVICE PTY LTD 36. Vertical input terminal for CH4 37. Trigger level LED 39. A time/div & delay time knob 40. B time/div knob 41. Variable adj of A sweep rate & x10 mag 42. Ready lamp Australia’s Largest Remarketer of Test & Measurement Equipment 20 Fulton Street, Oakleigh Sth, Vic., 3167. Tel: (03) 562 9500; Fax: (03) 562 9590 43. Calibration voltage terminals 44. Horizontal positioning of trace 45. Fine adjustment 46. Vertical input terminal for CH5 47. Delay time MULT switch 48. Selects between continuous & triggered delay 52. Trace separation adjustment 53. Ground terminal April 1995  75